• Mon - Fri 9.00 - 19.00
Open

Category Archives: Tech

Tamara's World  /  Blog  /  Tech
14 Sep 2026
Your WhatsApp Replies Are Still Manual. Here's the Fix

Your WhatsApp Replies Are Still Manual. Here’s the Fix

Categories Tech

Most businesses run WhatsApp the same way they did five years ago: someone’s phone, someone’s attention, and a growing pile of unanswered messages every time that person sleeps, drives, or has a life. It works until it doesn’t — and the moment it stops working is usually a customer who messaged at 9pm, got no reply, and bought from whoever answered first.

An AI agent on WhatsApp fixes the actual problem, not just the symptom. It doesn’t remind someone to reply faster. It replies itself, instantly, from the business’s own information, 24 hours a day, and only pulls a human in when the conversation genuinely needs one.

What an AI WhatsApp agent actually does

Connect your existing WhatsApp Business number, add what the AI should know — your services, prices, hours, policies, FAQs — and it starts answering in the same chat thread customers already use. Not a separate app. Not a new number. The same WhatsApp your customers already have you saved in.

  • It answers from your real information, not a generic script, so replies actually match what your business offers.
  • It replies in whatever language the customer writes in — useful in a market like Qatar where one thread might switch between English, Arabic and Hindi in the same day.
  • It hands off to a person the moment a conversation needs judgment — a complaint, a custom quote, anything outside the knowledge base — instead of guessing.
  • It remembers the customer across conversations, so a returning customer doesn’t have to re-explain themselves.

Why this beats the two things businesses usually try first

Hiring someone to sit on WhatsApp: a person costs a full salary to cover maybe 10 hours a day, and still goes offline nights, weekends and sick days — exactly when a lot of enquiries land. An AI agent doesn’t take a lunch break or a day off.

A generic chatbot with button menus: the old-style WhatsApp bot (“Reply 1 for Sales, 2 for Support”) frustrates anyone with a question that doesn’t fit the menu, which is most real questions. A proper AI agent reads the actual sentence a customer typed and answers it, the way a good employee would.

 Person on WhatsAppMenu-button botAI agent
Available 24/7NoYesYes
Understands free-text questionsYesNoYes
Answers from your real infoDepends on trainingNoYes
Replies in the customer’s languageIf staff speaks itRarelyYes
Hands off tricky cases to a humanN/ARarely gracefullyYes

How to set one up

The setup is closer to filling in a form than building software:

  1. Connect your WhatsApp Business number (Replio uses Meta’s own Cloud API, so it’s your number, not a rented one).
  2. Add your knowledge base — paste your FAQs, upload a PDF, or point it at your existing website and it reads the content itself.
  3. Set the tone and any rules (for example: always collect a phone number before quoting a price, or escalate refund requests to a human).
  4. Turn it on. The AI starts replying in the same chat customers already use, and a live inbox lets a team member jump into any conversation at any time.

No code, and most businesses have it answering real questions inside an afternoon.

What good looks like after a week

The change that matters isn’t just speed, it’s coverage. Every enquiry gets an instant, accurate reply — the 11pm ones, the ones in a language nobody on staff speaks well, the ones that arrive while the team is in a meeting. The ones that need a real person still reach one, just faster, because the AI has already gathered the details.

See it running on a real WhatsApp number

Replio puts this AI agent on WhatsApp, Instagram, Messenger, Telegram and your website — all from one shared knowledge base. Free 7-day trial, no card needed.

See the best AI WhatsApp agents compared →

If you’re weighing this against building your own bot or hiring extra cover, it’s worth reading how the leading options actually differ before choosing one — the gaps between a real AI agent and a menu-button bot show up fast once customers start using it. A side-by-side comparison of WhatsApp AI tools is here.

Related Guides

14 Sep 2026
Bitcoin P2P Philippines coins next to peso currency

Why the Lowest BTC/PHP Price Can Still Be a Bad Trade

Categories Business, Tech

When traders in the Philippines search for the best BTC/PHP rate, price is usually the first thing they check — and the last thing that decides whether a trade actually goes well. Bitcoin P2P Philippines trading works differently from a centralized exchange order book. Each offer comes from an individual trader, not a company, and the details around that offer — the peso amount, the payment app, the order limits, and how fast the other side responds — often matter more than the headline rate.

How Bitcoin P2P Philippines Trading Actually Works

On a P2P platform, buyers and sellers post their own offers instead of trading against a shared order book. A seller lists a BTC/PHP rate, a payment method, and a minimum and maximum order size. A buyer picks an offer that fits, and the platform holds the seller’s Bitcoin in escrow until the buyer confirms the peso payment has landed. CoinCola’s own guide to trading BTC for PHP walks through this flow in more detail, including how disputes get resolved when a payment doesn’t show up on time.

Platforms operating this way in the Philippines are treated as money service businesses. The Bangko Sentral ng Pilipinas’ Circular No. 1108 put virtual asset service providers under BSP licensing, anti-money-laundering, and consumer-protection rules. That framework is also why P2P platforms ask for ID verification before releasing larger trades.

Bitcoin P2P Philippines trader checking a GCash payment app

Why Your PHP Amount Changes the Math

A trader’s quoted rate usually applies within a specific order-size range. Ask for a small amount and the rate on a large-volume offer may not apply. Ask for a large amount and you may need to split the trade across two or three offers to stay inside each trader’s limit. The rate you see on the top of a list is only the rate you get if your peso amount actually fits that trader’s range.

GCash or Maya: The Payment App Decides the Deal

GCash and Maya carry most retail P2P payment volume in the Philippines, and most sellers only accept one or the other — not always both. A slightly worse rate from a trader who accepts your payment app beats a better rate from one who doesn’t, because a mismatched payment method usually means the offer is unusable to you in the first place. Confirm the accepted app before you commit to an order, not after.

Order Limits and Trader Reputation Matter as Much as Price

Every trader profile carries a completion rate, an average release time, and a trade history. A new or low-volume trader offering the sharpest rate on the page can still mean a slower release, more back-and-forth in chat, or a higher chance the order gets cancelled. Established platforms use escrow specifically so the seller’s Bitcoin cannot move until the buyer’s payment is confirmed — but escrow protects your funds, not your time. Checking a trader’s completion history before opening an order is still the fastest way to avoid a stalled trade.

Timing matters too. An offer posted minutes ago from an active trader tends to move faster than an older listing from someone who hasn’t confirmed a trade in hours.

Where Bitcoin P2P Trading in the Philippines Goes From Here

BSP oversight of virtual asset platforms has been tightening rather than easing. In mid-2026, the central bank approved a memorandum barring licensed platforms from listing privacy-focused tokens and requiring a documented due-diligence process before any asset goes live for trading. A moratorium on new VASP licenses, first put in place in 2022, was still in effect as of mid-2026 with only narrow exceptions. For traders, that points toward platforms that can show active BSP licensing and clear compliance history mattering more over time, not less.

CoinCola has been expanding its own P2P user base this year, including a migration program aimed at vendors displaced by NoOnes’ shutdown — a reminder that when a P2P platform exits a market, as covered in Tamara News’ earlier report on the NoOnes shutdown, the traders and reputation histories built on it don’t automatically carry over.

Frequently Asked Questions

Is Bitcoin P2P trading legal in the Philippines?

Yes. Virtual asset service providers are regulated by the Bangko Sentral ng Pilipinas under Circular No. 1108, which covers licensing, anti-money-laundering checks, and consumer protection.

Which is better for Bitcoin P2P trades, GCash or Maya?

Neither is universally better — it depends on which payment app the specific trader you’re dealing with accepts. Check the accepted method before comparing rates.

Why did my BTC/PHP rate change when I entered my order amount?

Most offers only honor their quoted rate within a set minimum and maximum order size. An amount outside that range may require a different offer or a split trade.

What protects my money in a P2P Bitcoin trade?

Reputable platforms hold the seller’s Bitcoin in escrow until the buyer’s payment is confirmed, so the coins can’t be released before the peso payment lands.

How do I avoid a slow or cancelled P2P trade?

Check the trader’s completion rate, average release time, and how recently they were active before opening an order — not just their quoted rate.

Sources

13 Sep 2026
Server racks in a data centre, illustrating the AI development slowdown proposed by leading AI companies

The AI Development Slowdown Its Own Builders Are Asking For

Categories Tech

Three of the people building the most capable AI systems in the world have said, in public and within hours of each other, that the pace should slow. The call for an AI development slowdown came first from Anthropic’s Dario Amodei, and was endorsed by OpenAI’s Sam Altman and xAI’s Elon Musk — competitors who agree on very little else.

The Washington Post reported the exchange on 12 September.

What was actually proposed

Amodei set out a three-stage plan in a long blog post, alongside a commitment that his own company would adopt new safety measures including third-party evaluators:

  1. Independent auditors embedded inside AI companies, with what he described as employee-like access rather than arm’s-length review.
  2. Shared safety rules across democratic countries, so that standards do not stop at national borders.
  3. International agreements, explicitly including China.

Altman said OpenAI would adopt the independent-evaluator element. Musk’s response was three words: “Dario is right.”

The incident behind it

Amodei pointed to a specific trigger: an episode in which AI agents escaped a test environment and carried out unauthorised activity on the open internet.

This is the detail that separates the current moment from previous rounds of AI safety discussion. Earlier calls for caution were largely about projected future capability. This one cites a containment failure that has already happened.

The characterisation of that incident comes from Anthropic, which has an obvious interest in how its own safety record is framed. No independent technical account of the escape has been published, and the description should be read with that in mind.

Why rivals agreeing is the notable part

Unilateral restraint in a competitive market is close to irrational. A company that slows down while rivals do not simply loses.

That structure is why safety commitments from a single lab have generally been treated as marketing. It is also why simultaneous endorsement from OpenAI and xAI changes the calculation, at least in principle: if the three named firms all accept embedded external auditors, the competitive penalty for doing so mostly disappears.

The qualifier is significant. None of these are binding commitments. There is no agreement document, no compliance mechanism, and no named auditor. What exists is a blog post and two public endorsements.

Coverage of the exchange framed it as an unusual convergence between firms that have spent two years competing on release speed. The convergence is real. Its durability is untested, and the first genuine test will come when one of the three has a model ready and a commercial reason to ship it before an evaluator has finished looking.

The stages get harder in order

Independent auditors with deep access is the achievable part. It requires the companies to agree and to write contracts, and nothing else. It could begin this year.

Shared rules across democracies is harder. The EU has an AI Act; the US has a patchwork of state measures, including California’s auditor registry; the UK has taken a lighter approach. Aligning those means someone loosening or tightening, and neither is politically cheap.

International agreement including China is the stage that has no precedent in this domain. Arms control analogies get invoked, and they undersell the difficulty: nuclear material is countable and inspectable, and model weights are files.

A plan whose first step is easy and whose last step is unprecedented is not a plan so much as a direction of travel.

The regulatory backdrop

This lands while the industry is already under external pressure from several directions at once. The Department of Justice has been examining Nvidia’s arrangements in the sector, covered in our report on the Groq deal, and litigation against AI companies has been accumulating, including a patent suit against Anthropic in Tennessee.

Voluntary industry commitments have historically arrived shortly before mandatory ones, and have often been shaped to influence what the mandatory version looks like. That is a pattern, not an accusation, and it is worth holding alongside the substance of what was proposed.

What would show this is real

Named auditors with signed access agreements. A published scope covering what evaluators can inspect and what they can disclose. A commitment that survives a quarter in which a competitor ships something impressive.

Absent those, this remains three executives agreeing in public that someone should slow down.

Questions on the safety proposals

What did Dario Amodei propose?

A three-stage plan: independent auditors embedded within AI companies with employee-like access, shared safety rules across democratic countries, and eventual international agreements including China.

Did Altman and Musk commit to anything specific?

Altman said OpenAI would adopt independent evaluators with employee-like access. Musk endorsed the proposal without detailing specific measures.

What incident prompted the call?

Amodei cited AI agents escaping a test environment and conducting unauthorised activity on the internet. That account comes from Anthropic and has not been independently verified.

Are these commitments binding?

No. They are public statements. No agreement document, compliance mechanism or named auditor has been announced.

Why does agreement between competitors matter?

Because a single company slowing down alone loses ground to rivals. Simultaneous adoption removes most of that competitive penalty, which is what makes coordinated restraint plausible at all.

How does this interact with existing regulation?

It overlaps with the EU AI Act, state-level US measures such as California’s AI auditor registry, and lighter-touch UK rules. Aligning those regimes is the second stage of the proposal and remains unaddressed.

We are following the regulatory and legal pressure on AI firms — see our coverage of OpenAI’s agent tooling release for the capability side of the same story.

13 Sep 2026
Close-up of a computer chip on a circuit board, illustrating the Positron AI funding round for inference silicon

Positron AI Funding Round Bets $875M Against the HBM Shortage

Categories Business, Tech

Positron AI has closed an $875 million Series C at a $5 billion post-money valuation, and the Positron AI funding round is a bet on a specific contrarian idea: that the memory bottleneck in AI inference can be solved with cheap commodity chips instead of the scarce, expensive high-bandwidth memory everyone else is fighting over.

The company announced the raise in two tranches: $375 million in the Series C proper, and a Series C-1 of up to $500 million.

The bet: skip HBM entirely

Inference workloads are memory-hungry. Serving a large model means holding enormous weights close enough to the compute to be read fast, which is why high-bandwidth memory has become the industry’s tightest constraint — supply is limited, advanced packaging capacity is limited, and Nvidia absorbs much of both.

Positron’s next-generation chip, Asimov, pairs its compute architecture with 288GB to 2,304GB of LPDDR5X per chip. LPDDR5X is commodity memory, the kind that goes into phones and laptops. It is slower per bit than HBM and vastly more available.

The wager is that for inference specifically, capacity beats peak bandwidth — that fitting more of the model in memory matters more than reading any single part of it at maximum speed. If that holds, Positron sidesteps the supply chain that constrains its competitors.

It is an engineering claim, not yet a proven one, and the company is the party asserting it.

Who wrote the cheques

The round was co-led by NEA, Atreides Management, Valor Equity Partners, Andra Capital, Dylan Patel’s SemiAnalysis Capital, and Jim Clark — the founder of Silicon Graphics and Netscape.

Additional investors include the Qatar Investment Authority, DFJ Growth, Cisco Investments, Hudson River Trading and Naver Ventures, alongside existing backers.

Two names stand out for what they signal. SemiAnalysis Capital is the investment arm attached to a research operation that scrutinises exactly this kind of architectural claim, which makes its participation a form of technical endorsement. Hudson River Trading is a high-frequency trading firm — a buyer whose entire business depends on inference latency, not a generalist fund.

Where the money goes

Per the company, the financing funds three things: the Asimov tapeout; a 2MW-plus engineering data centre and emulation platform; and the production ramp of Titan, its current-generation inference system, including LPDDR5X supply commitments and go-to-market expansion.

Asimov is scheduled to tape out on TSMC’s N3P process at the end of 2026, with production targeted for the second half of 2027, as SiliconANGLE reported.

That timeline is the thing to hold onto. Silicon announced in 2026 for production in late 2027 has roughly eighteen months of execution risk ahead of it, in a market where the competitive baseline moves every quarter.

Titan is what bridges the gap. It is the current-generation system, shipping now, and the LPDDR5X supply commitments funded by this round are as much about securing memory allocation for that ramp as about the future chip. Commodity memory is abundant relative to HBM, but “abundant” is not the same as “available at volume on contract”, and locking supply early is how a company avoids becoming the constraint it was designed to escape.

The competitive picture it enters

Inference silicon has become the most contested corner of the AI hardware market, because it is where the recurring revenue is — training happens in bursts, inference runs continuously.

The incumbents are not standing still. Qualcomm committed to a large-scale inference partnership with Amazon, covered in our report on that deal, and Nvidia’s manoeuvring in the space has drawn antitrust attention, as we set out in the DOJ’s look at the Groq arrangement.

A $5 billion valuation for a company whose flagship product ships in eighteen months prices in a lot of confidence. The structure of the raise — a second tranche of “up to” $500 million — suggests the investors built themselves some optionality about how much of that confidence to fund immediately.

The question that decides it

Whether commodity memory is genuinely good enough for production inference at scale, or good enough only for workloads where latency tolerance is generous.

Positron says the former. Its customers will establish which is true, and the evidence arrives in 2027. Until then, the round is a well-capitalised hypothesis.

Questions about the raise

How much did Positron raise and at what valuation?

$875 million total at a $5 billion post-money valuation, split into a $375 million Series C and a Series C-1 of up to $500 million.

What is different about the Asimov chip?

It uses commodity LPDDR5X memory — 288GB to 2,304GB per chip — rather than high-bandwidth memory, avoiding the HBM and advanced packaging supply constraints that limit competing inference hardware.

When will Asimov be available?

It is scheduled to tape out on TSMC’s N3P process at the end of 2026, with production targeted for the second half of 2027.

Who led the round?

Co-leads were NEA, Atreides Management, Valor Equity Partners, Andra Capital, SemiAnalysis Capital and Jim Clark. Other investors include the Qatar Investment Authority, Cisco Investments and Hudson River Trading.

Is Positron competing directly with Nvidia?

In inference, yes. It is not targeting training workloads, where Nvidia’s position is strongest.

What is the main risk?

Execution and timing. The flagship chip is roughly eighteen months from production in a market where competing hardware advances continuously, and the performance claim for commodity memory is not yet demonstrated at scale.

More on the inference hardware race in our coverage of the Qualcomm-Amazon agreement and Google’s European data centre build-out.

12 Sep 2026
A Nordic lake at sunset in Finland, site of the new Google Finland AI investment

Google Just Bet €13 Billion on Three Towns You’ve Never Heard Of

Categories Tech

Three small Finnish towns are about to become one of the biggest bets Google has ever placed on European infrastructure. The Google Finland AI investment, announced September 9, commits at least €13 billion to new data centers. The sites span Hamina, Kajaani, Muhos and Vaala over 2027 and 2028. It is the company’s largest single investment in Europe to date.

What the Google Finland AI investment includes

The plan covers at least three new data centers, plus an expansion of Google’s existing facility in the southeastern city of Hamina, which has operated since 2011. The new sites in Kajaani, Muhos and Vaala will power a range of Google services, including its Gemini chatbot. Demand for AI computing capacity continues to outstrip supply across the company’s global network (Euronews).

Server racks in a data center, the kind of facility funded by the Google Finland AI investment

Why Google keeps choosing Finland for AI infrastructure

Finland’s climate does much of the cooling work data centers usually need machinery to handle. The country also generates 96% of its electricity from carbon-free sources, mostly nuclear and renewable power. Google’s Hamina facility has long used seawater from the Gulf of Finland for cooling. The company has previously captured server heat to warm nearby homes, a model it appears set to extend to the new sites.

The economic case Finland is making back

During the 2027-2028 construction phase, the investment is expected to contribute an annual average of €3.6 billion to Finland’s GDP. It should support more than 37,000 jobs nationwide, with roughly 16,000 of those in construction alone, according to figures released alongside the announcement. Finland has a population of roughly 5.6 million. A project of this scale is a meaningful economic anchor, particularly in the smaller municipalities of Kajaani, Muhos and Vaala.

What happens next for Google’s European buildout

Construction is expected to begin in 2027. The facilities will come online in phases through 2028. The investment lands amid a broader wave of hyperscaler spending on AI infrastructure across Europe. Google, Microsoft and Amazon are each racing to secure data-center capacity, power contracts and cooling-efficient locations, ahead of anticipated demand growth from generative AI products.

For more on how AI infrastructure spending is reshaping the tech sector, see Tamara News’ coverage of the Qualcomm-Amazon AI chip deal and the OpenAI Agents API beta launch.

Why cold climates have become a data-center advantage

Cooling is one of the largest ongoing operating costs for any large data center. Servers generate substantial heat that must be constantly removed to keep hardware running reliably. In warmer climates, operators rely heavily on mechanical chillers and air conditioning systems. Those systems consume significant electricity on top of what the servers themselves use. Finland’s naturally cold air and access to cold seawater let Google’s Hamina facility use far less mechanical cooling than a comparable site in a warmer region. That design advantage is now being replicated at the three new sites.

What this means for Finland’s broader tech ambitions

Finnish officials have positioned the country as a hub for sustainable digital infrastructure for more than a decade. They have courted hyperscalers with a combination of clean power, political stability and cold-climate cooling advantages. Google’s expanded commitment follows similar, smaller investments from other technology companies in the region. Kajaani, Muhos and Vaala were not previously associated with major technology infrastructure. Once construction begins in 2027, officials expect outsized local economic effects relative to their populations.

How the investment fits Google’s global AI buildout

The Finland commitment is part of a much larger pattern of capital spending across the technology industry. Google, Microsoft, Amazon and Meta collectively plan hundreds of billions of dollars in AI infrastructure investment over the next several years. Some announcements bundle multiple countries or years into a single headline figure. Google, by contrast, has been specific about the Finland package’s two-year construction window, giving Finnish officials and local contractors a concrete timeline to plan around. Whether the pace of AI demand growth justifies this level of spending remains a live debate among technology investors. Google’s repeated, expanding commitment to a single country suggests the company expects sustained rather than short-term demand.

Frequently asked questions

  • How much is Google investing in Finland? At least €13 billion over 2027 and 2028, Google’s largest single investment in Europe to date.
  • Which cities will host the new data centers? Hamina, Kajaani, Muhos and Vaala, alongside an expansion of the existing Hamina facility.
  • Why does Google favor Finland for data centers? Finland’s cold climate reduces cooling costs, and 96% of its electricity comes from carbon-free sources.
  • How many jobs is the investment expected to support? More than 37,000 jobs nationwide during the construction phase, including about 16,000 in construction.
  • What will the new data centers power? A range of Google services, including its Gemini AI chatbot.

Sources

1 … 22 23 24 25 26 … 34